Concrete, as the predominant construction material, plays a vital role in human development. However, its production involves a complex mix of factors, such as the presence of raw materials, hydration reactions, curing conditions, and environmental parameters resulting in uncertainties and variations in the final material properties. Achieving desired concrete properties necessitates a deep understanding of these interdependencies and the ability to optimize the mixture accordingly. Furthermore, the growing demand for sustainable and environmentally friendly concrete adds an additional layer of complexity to the production process. Machine learning (ML) has emerged as a transformative tool in concrete research, offering autonomous solutions to complex tasks. To the best of our knowledge, this is one of the first studies to tackle concrete mix design using ML, making the proposed research a frontier one. Different machine learning models were trained on a dataset of 1100 experimentally tested concrete samples to define the consistency class and the compressive strength of concrete mixtures on the basis of the mix design components. The results indicated that the proposed approach is well suited for the former task (~80% accuracy), while the latter scenario still requires deeper data analysis and exploration. In any case, the results obtained demonstrate that the presented preliminary research is highly promising, as the achieved results show a level of accuracy comparable to that obtained through the current human-based procedure actually adopted in concrete casting operations.